Examples disclosed herein relate to an autonomous driving system in an ego vehicle. The autonomous driving system includes a radar system configured to detect and identify a target in a path and a surrounding environment of the ego vehicle. The autonomous driving system also includes a sensor fusion module configured to receive radar data on the identified target from the radar system and compare the identified target with one or more targets identified by a plurality of perception sensors that are geographically disparate from the radar system. Other examples disclosed herein include a method of operating the radar system in the autonomous driving system of the ego vehicle.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
2. The autonomous driving system of claim 1, wherein the radar system comprises a metamaterial antenna structure configured to radiate the one or more transmission RF beams and receive one or more return RF beams reflected from the surrounding environment, wherein the sensor fusion module is configured to send a control signal to the metamaterial antenna structure based on historical sensor data from the radar system, and wherein the control signal enables one or more metamaterial antenna cells in the metamaterial antenna structure to be directed.
3. The autonomous driving system of claim 2, wherein the control signal comprises an instruction to the metamaterial antenna structure to radiate additional transmission RF beams at a given phase shift and direction within at least a portion of a field-of-view corresponding to a location of a target identified by the radar system.
4. The autonomous driving system of claim 2, wherein the radar system comprises a perception module coupled to the metamaterial antenna structure, and wherein the perception module is configured to generate tracking information of an identified target with a multi-object tracker in the perception module.
5. The autonomous driving system of claim 4, wherein the multi-object tracker is configured to track the identified target over time using a Kalman filter.
6. The autonomous driving system of claim 4, wherein the perception module is further configured to generate target identification information based at least on the tracking information.
7. The autonomous driving system of claim 6, wherein the radar system is further configured to combine the target identification information with other target identification information from the plurality of perception sensors to form the combined target identification information.
The autonomous driving system is designed to enhance object detection and identification for self-driving vehicles. The system addresses the challenge of accurately identifying and tracking objects in dynamic driving environments by integrating data from multiple perception sensors, including radar. The radar system is configured to process target identification information, such as object location, velocity, and classification, and then combine this data with information from other sensors, such as cameras or lidar, to form a more comprehensive and reliable set of target identification data. This fusion of sensor data improves the system's ability to detect and track objects under varying conditions, reducing false positives and enhancing overall situational awareness. The combined target identification information is used to support decision-making processes, such as path planning and collision avoidance, ensuring safer and more efficient autonomous driving operations. The system leverages advanced signal processing and sensor fusion techniques to optimize object detection performance, particularly in scenarios where individual sensors may have limitations, such as occlusions or adverse weather conditions.
8. The autonomous driving system of claim 7, wherein the radar system is further configured to send the combined target identification information to the sensor fusion module in the autonomous driving system.
9. The autonomous driving system of claim 7, wherein the sensor fusion module is further configured to receive the other target identification information over a vehicle-to-vehicle communication channel from the plurality of perception sensors.
10. The autonomous driving system of claim 7, wherein the sensor fusion module is further configured to generate enhanced target identification information from the combined target identification information, the enhanced target identification information including one or more adjustments to the identified target in terms of time and position relative to the ego vehicle.
Autonomous driving systems rely on sensor fusion to combine data from multiple sensors (e.g., cameras, radar, LiDAR) to identify and track targets (e.g., vehicles, pedestrians) in the environment. A key challenge is accurately determining the position and movement of these targets relative to the ego vehicle (the autonomous vehicle itself) to enable safe navigation. Existing systems may struggle with inconsistencies or errors in sensor data, leading to misidentification or inaccurate tracking of targets. This invention improves sensor fusion in autonomous driving by generating enhanced target identification information. The system combines raw target identification data from multiple sensors into a unified set of target data. It then processes this combined data to refine the identified targets, adjusting their time and position relative to the ego vehicle. These adjustments account for sensor discrepancies, environmental factors, or dynamic changes in the target's movement, resulting in more accurate and reliable target tracking. The enhanced information helps the autonomous vehicle make better decisions, such as adjusting speed, steering, or braking to avoid collisions or navigate complex scenarios. This approach enhances the robustness and safety of autonomous driving systems by improving the accuracy of target detection and tracking.
11. The autonomous driving system of claim 10, wherein the sensor fusion module is further configured to determine a next control action for the metamaterial antenna structure based at least on the enhanced target identification information.
14. The radar system of claim 13, wherein the perception module is further configured to generate tracking information of the identified target with a multi-object tracker in the perception module.
A radar system is designed to detect and track objects in an environment, addressing challenges in accurately identifying and monitoring multiple moving targets. The system includes a perception module that processes radar data to identify targets and generate tracking information. This tracking information is produced using a multi-object tracker within the perception module, which enables the system to maintain and update the positions, velocities, and identities of multiple targets over time. The multi-object tracker improves situational awareness by distinguishing between separate objects, reducing false detections, and providing reliable trajectory data. This capability is particularly useful in applications such as autonomous vehicles, surveillance, and collision avoidance, where precise and continuous tracking of multiple objects is essential for safety and decision-making. The system may also include additional components, such as a sensor fusion module that integrates data from multiple sensors to enhance target detection and tracking accuracy. By combining radar data with inputs from other sensors, the system achieves robust performance in various environmental conditions. The overall design ensures reliable object tracking, even in dynamic and cluttered scenarios.
15. The radar system of claim 14, wherein the multi-object tracker is configured to compare one or more candidate targets identified by the target identification and decision module with targets that the multi-object tracker has detected in one or more prior segments of time.
18. The method of claim 17, wherein the target identification information comprises one or more of a classification of the identified one or more targets, a location of the identified one or more targets, or a rate of movement of the identified one or more targets.
20. The method of claim 17, wherein the enhanced target identification information provided by the sensor fusion module is used in training one or more deep learning networks of the perception module.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
June 5, 2019
October 25, 2022
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.